In order to investigate and correctly measure the monetary benefits obtained when human capital migrates, we have to control for individuals’ characteristics and as many aspects of their “life history” as possible, regardless of the selected statistical approach. Individuals’ backgrounds and experiences will affect their propensity to migrate; as such, only by comparing similar individuals is it possible to estimate the monetary effect of choosing to work abroad after graduation. Socio-economic backgrounds, training choices, learning experiences, and more general life stories all contribute to the propensity of migrating, as well as the possible outcomes. This is a well-known bias problem generated by a self-selection mechanism. We present a method using a statistical adjustment of the self-selection problem applied to a database of life history records. The AlmaLaurea database used in this paper collects thorough data from graduates of Italian universities, including information on graduates’ previous studies and social backgrounds. We analyse this data to measure the impact of the emigration choice in the context of human capital theory as applied to the modern view of a global labour market. We perform a statistical multivariate analysis to study migrants vs. non-migrants, using an innovative data-driven potential outcome approach that corrects for self-selection bias. We continue with a cluster analysis to identify homogeneous groups based on a multiple correspondence analysis and testing the self-selection bias reduction using a global imbalance measure.

Camillo, F., Benassi, S., Vittadini, G. (2019). ITALIAN GRADUATES AND INTERNATIONAL MOBILITY: A POTENTIAL OUTCOME MODEL APPLIED TO ALMALAUREA DATA. STATISTICA APPLICATA, 31(1), 157-176.

ITALIAN GRADUATES AND INTERNATIONAL MOBILITY: A POTENTIAL OUTCOME MODEL APPLIED TO ALMALAUREA DATA

Vittadini, G.
2019

Abstract

In order to investigate and correctly measure the monetary benefits obtained when human capital migrates, we have to control for individuals’ characteristics and as many aspects of their “life history” as possible, regardless of the selected statistical approach. Individuals’ backgrounds and experiences will affect their propensity to migrate; as such, only by comparing similar individuals is it possible to estimate the monetary effect of choosing to work abroad after graduation. Socio-economic backgrounds, training choices, learning experiences, and more general life stories all contribute to the propensity of migrating, as well as the possible outcomes. This is a well-known bias problem generated by a self-selection mechanism. We present a method using a statistical adjustment of the self-selection problem applied to a database of life history records. The AlmaLaurea database used in this paper collects thorough data from graduates of Italian universities, including information on graduates’ previous studies and social backgrounds. We analyse this data to measure the impact of the emigration choice in the context of human capital theory as applied to the modern view of a global labour market. We perform a statistical multivariate analysis to study migrants vs. non-migrants, using an innovative data-driven potential outcome approach that corrects for self-selection bias. We continue with a cluster analysis to identify homogeneous groups based on a multiple correspondence analysis and testing the self-selection bias reduction using a global imbalance measure.
Articolo in rivista - Articolo scientifico
Human capital theory, migration, higher skilled, selection bias treatment, wage premium
English
1-apr-2019
2019
31
1
157
176
open
Camillo, F., Benassi, S., Vittadini, G. (2019). ITALIAN GRADUATES AND INTERNATIONAL MOBILITY: A POTENTIAL OUTCOME MODEL APPLIED TO ALMALAUREA DATA. STATISTICA APPLICATA, 31(1), 157-176.
File in questo prodotto:
File Dimensione Formato  
Italian Graduates and International Mobility A Potential Outcome Model Aplied to Almalaurea Data.pdf

accesso aperto

Descrizione: articolo principale
Tipologia di allegato: Publisher’s Version (Version of Record, VoR)
Dimensione 4.64 MB
Formato Adobe PDF
4.64 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/279622
Citazioni
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
Social impact